Live Media Streaming Latency Reduction via Segment Polling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for streaming live media, such as MPEG-DASH, experience significant latency variation between client requests and playback due to large segment durations and clock synchronization issues, leading to desynchronization among clients, particularly in live events like sports coverage.

Innovation Solution

A method where a client determines the next available segment using a manifest and starts polling for it once acknowledged, reducing latency by initiating playback as soon as the segment is published, with a timeout mechanism to prevent indefinite waiting.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If large segment durations are used for media streaming, then network transmission efficiency is improved, but latency variation between client requests and playback increases

Engineering Contradiction:

Improvenetwork transmission efficiencyVSAvoidlatency variation

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The client proactively determines the next available segment using manifest information before actually requesting it. By calculating which segment will be available next based on manifest timing data, the client prepares the request in advance, reducing the latency variation that would otherwise occur due to large segment durations.

Inventive Principle:

Principle #10Preliminary action

2Device complexity

If clients use local clocks to determine segment availability, then device complexity is reduced, but synchronization accuracy among clients deteriorates

Engineering Contradiction:

Improveclient device complexityVSAvoidclock synchronization accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The manifest file serves as an intermediary that provides authoritative timing information from the server. Instead of relying on potentially unsynchronized local client clocks, the manifest contains precise timing data that all clients use to determine segment availability, thereby achieving synchronization without increasing device complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Reliability

If clients wait for segment availability confirmation before playback, then playback reliability is improved, but latency increases

Engineering Contradiction:

Improveplayback reliabilityVSAvoidend-to-end delay

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The client uses manifest information to determine the next available segment and initiates playback based on this feedback from the server's published timing data. This feedback mechanism allows the client to start playback as soon as the segment is published without waiting for additional confirmation, reducing end-to-end delay while maintaining reliability through the manifest's authoritative timing information.

Inventive Principle:

Principle #23Feedback

Data Source

PatentEP3318034B1Low latency media streaming

Publication Date: 2023.10.25 BRITISH TELECOM PLC

Figure 1
Figure 2
Figure 3

AI summary

Examples of the present invention provide a method of reducing the latency in streaming live media by a client from a server. The client uses manifest information to determine the "live edge" of the live media stream, where the live edge is represented by the segment from the media stream corresponding to the current time. The client then uses this to identify the next segment, which is the segment in time that will next become available. The client then starts making repeated polling requests for that next segment until the segment becomes available. As a result, the newest possible segment is obtained by the client as soon as it becomes available, and latency is reduced. Further, when adopted by all clients, the latency variation between clients is also reduced.